Solving the congestion problem of Redis cluster Jedis drive under high concurrency

After Redis Cluster was released, we used cluster in the project, and the driver was Jedis, but during the stress test it was found that a certain number of Redis accesses were very slow up to dozens of seconds and a few minutes, After analysis Jedis drive Jedisclusterinfocache in the lock caused by

Private map<string, jedispool> nodes = new hashmap<string, jedispool> ();
Private Map<integer, jedispool> slots = new Hashmap<integer, jedispool> ();

Public Jedispool getnode (String nodekey) {
R.lock ();
try {
Return Nodes.get (Nodekey);
} finally {
R.unlock ();
}
}

Public Jedispool getslotpool (int slot) {
R.lock ();
try {
return Slots.get (slot);
} finally {
R.unlock ();
}
}

Replace the map with Concurrenthashmap

Private map<string, jedispool> nodes = new concurrenthashmap<string, jedispool> ();
Private Map<integer, jedispool> slots = new Concurrenthashmap<integer, jedispool> ();

Remove lock

Public Jedispool getnode (String nodekey) {
Return Nodes.get (Nodekey);
}

Public Jedispool getslotpool (int slot) {
return Slots.get (slot);
}

After testing the comparison process is relatively smooth, there is no time for a visit to a very long situation.

This article is from "Tianya Time and Space" blog, please make sure to keep this source http://leshjmail.blog.51cto.com/629172/1784716

Solving the congestion problem of Redis cluster Jedis drive under high concurrency